measuring characteristic impedance

last updated 25 March 2025.

Usually when we buy coax cable or other feedline, we get a well-known type with a specified characteristic impedance – often 50 ohms, and we don’t give it much thought.

Occasionally, however, we may encounter an unmarked piece of cable, or one that we don’t have data for, and want to determine the impedance to see how we might use it. There are several different ways to measure the characteristic impedance, depending on what test equipment you have available.

First, let me say that you should consider all of these methods approximate, especially with equipment design for ham use rather than laboratory precision. But also, characteristic impedance is not a fixed value – it actually varies with frequency. Not necessarily by a lot, but by measurable amounts. For example, one nominal “50 ohm” cable that I was using actually was 54 ohms at 1 MHz, 52 ohms at 4 MHz, 50.75 ohms at 30 MHz, and dropping closer to 50 ohms as the frequency approached 1 GHz. So the same cable may measure differently depend on what frequency you use to test it.

And the characteristic impedance isn’t even a real value, but a complex number, like 50.5 – j0.5 ohms (for RG-58 coax at 30 MHz). (The reactive term accounts for the cable losses, but I won’t try to explain that here.) So understand that we are just finding a value that is reasonably close enough for us to know what to expect with the feedline with regards to matching, but not necessarily precision.

50 or 75 ohms?

Often we just want to determine whether it is 50 or 75 ohm coax, the two most common values. In that case, we just have to be able to tell the two apart. (Other values for RG-type coaxial cables are 25, 93, or 125 ohms.) These are three of the easiest methods:

method #1: capacitance meter

This test requires only a short piece of cable. For types with solid polyethylene insulation, the capacitance between the center conductor and the shield is around 100 pF per meter (30 pF per foot) for 50 ohm cables, and 67 pF per meter (20 pF / foot) for 75 ohm cables. Values are lower for types with foam dielectric: down to perhaps 75 pF/m (23 pF/ft) for very low loss 50 ohm cables and 53 pF/m (16 pF/ft) for 75 ohm cables. With a digital capacitance meter, even 1m (3 feet) of cable is sufficient, especially for solid dielectric cables (otherwise there may be some ambiguity in the 67 – 75 pF/m (20 to 23 pF/ft) range. Just measure the capacitance and divide by the cable length.

method #2: check the SWR

Put a 50 ohm dummy load on the end of the cable and connect it to the transmitter through an SWR meter. If the SWR reads over about 1.5 : 1 (assuming a good dummy load), then it probably is 75 ohm cable. If it measures close to 1 : 1 on all bands, it is 50 ohm cable. Well, as long as the cable is long enough – preferably at least 1/4 wavelength.

The impedance at the radio will vary from 50 ohms to 112 ohms when a 75 ohm cable is terminated with a 50 ohm load. When the cable is an odd multiple of 1/4 wavelength, the impedance will be over 100 ohms, and the SWR will be greater than 2 : 1. Fortunately this length doesn’t need to be too accurate for the test to work, so you can cut a quarter wave using an estimated velocity factor of 0.7 to 0.75 or so, and you should still see enough difference. When the cable is a multiple of 1/2 wavelength, the impedance at the transmitter will be close to 50 ohms regardless of the cable impedance, which is why you may need to test it on various bands to be sure.

If you have an antenna analyzer, you can scan the SWR across an range of frequencies. With a 50 ohm load on a 75 ohm cable, you should see the SWR rising to around 2 : 1 on some frequencies and dropping to 1 : 1 on others.

method #3: measure the SWR of a long length

When one end of a cable is open or shorted, then the impedance at the other end of the cable will approach the characteristic impedance when the cable is long enough to attenuate the reflected wave. This is more convenient at the higher frequencies, VHF or UHF, unless you have a large spool, as it may require 10 – 20 wavelengths to get a reliable result. If you have analyzer, you can scan a frequency range that includes several minima and maxima in the R value, and eyeball average the center line of the plot. One nice thing about this approach is that you can test a large spool of cable without unwinding it, or even knowing whether the far end is open or shorted. When the line has 10 dB of loss, the impedance at the transmitter should vary between about 40 and 60 ohms for a 50 ohm cable (SWR of 1.2 : 1 or less).

other impedances

Sometimes we don’t know what impedance to expect, especially we are using some sort of wire that wasn’t intended for RF use, like speaker cable or AC power cord. Then we want a more accurate estimate of the characteristic impedance, and we have less of an idea what range to expect. This is easier with an antenna analyzer, but some tests can be made with simpler equipment.

Method #1 doesn’t work as well in this case, as the range of possible capacitance values doesn’t give enough resolution to identify the characteristic impedance with enough detail. It might give you an idea, however, especially for solid polyethylene dielectric cables.

method #2: check the SWR – extended

This method can be extended to other impedance cables. When terminated with a 50 ohm dummy load (or other known load impedance, R_load), the impedance at the transmitter end of a 1/4 wavelength cable (R_measured) will be given by:

R_measured = Z₀ * Z₀ / R_load

where Z₀ is the characteristic impedance of the line (at least, if we ignore losses in the feedline).

so,

Z₀ = the square root of ( R_load * R_measured )

Usually the characteristic impedance will be equal to or greater than 50 ohms, in which case an SWR measurement may be close enough ( as long as your line is close to 1/4 wavelength ). So an SWR of 4 : 1 with a 50 ohm dummy load would imply a measured impedance of 200 ohms, and the calculated Z₀ would be 100 ohms. Many analog SWR meters have limited accuracy over 3 : 1 or so, and that may limit the use of this approach unless a digital meter or an analyzer is used.

Note that an SWR of 4 : 1 could also be caused by a Z₀ of 25 ohms: such coax cables aren’t very common in ham use, but the do exist. To resolve the ambiguity, replace the dummy load with a 100 ohm resistor. The SWR at the transmitter would be 2 : 1 for a 100 ohm cable, or; 8 : 1 for a 25 ohm cable.

Of course, using an analyzer, you can simply plot R and X vs. frequency, take the lowest frequency where SWR is maximum, and look at the R value.

If you have a handful of non-inductive resistors (such as carbon composition types – can check them on the analyzer) you can try different ones and see which one gives the flattest SWR curve vs. frequency across the HF band. You can also reverse the previous equation and search for a load resistor that gives a low value of SWR₅₀ at the analyzer, and calculate the characteristic impedance from that. I often use a combination of methods to come up with an estimate, especially when the line has relatively high losses.

method #3: measure the SWR of a long length – extended

This may work for coax cables or other shielded lines, but for parallel-conductor lines that aren’t shielded you probably need to unwind the spool and stretch the cable across the garden or along a wood fence to get a good reading. It isn’t as convenient.

method #4: the reactance of a 1/8 wavelength cable

With the far end of the feedline open, find the lowest frequency where the resistance is low and the reactance is zero: that should be the frequency where the feedline is 1/4 wavelength (electrically – a bit shorter than the calculated length of a quarter wave wire). Now set the analyzer to exactly half the frequency and measure the complex impedance: the value of the reactance should be the characteristic impedance of the feedline.

In theory this also works if the far end of the cable is shorted, in which case the 1/4 wavelength point is the lowest frequency where the R value is high. But my experience is that, if I’m using any sort of adaptors to connect the coax to my analyzer (because the cable doesn’t have a proper connector on it), then that has more effect on the reading.

example

Let’s use an example the same 1.75m (5.75 foot) piece of parallel-conductor power cable that we used for measuring the velocity factor.

We’ll start with method #4. With my analyzer, I measured the quarter wave resonant frequency as 27.9 MHz (with the far end open). So I set the analyzer to half that value (13.95 MHz) and looked at the reactance: -127.3 ohms. So we guess that the impedance is around 127 ohms.

Note, however, that some parts of this measurement depend on the measurement accuracy and how the cable is connected to the analyzer. Initially I had used two adaptors between the cable and the analyzer, and the resonant frequency was 29.3 MHz. I ended up having to connect the cable directly to the coax jack on the analyzer (using tape to hold one conductor to the outside of the jack, and the other stuck into the center pin) to get more accurate results. That’s less of an issue when measuring longer cables.

This method should work with the far end shorted as well. When I tried that, the R value jumped around quite a bit, while the X value was relatively stable at about +134 ohms.

So 130 ohms is a reasonable working number. But let’s confirm that using method #2.

If the impedance is 130 ohms, then a load resistor of 130 * 130 / 50 = 338 ohms should give a good 50 ohm match at the feedpoint.

330 ohms is a standard resistor value, so I grabbed a 1/4 watt resistor from my junkbox. My analyzer read 55 ohms, which would suggest the impedance is closer to 135 ohms instead. But the standard formulas for impedance transformation ignore the cable losses, so that isn’t unusual.

The next step was to find a 130 ohm resistor (a standard 10% value between 120 and 150 ohms) and look at the SWR curve vs. frequency: we would expect pretty close to a constant ( 130 / 50 = ) 2.6 : 1 across a large part of the HF spectrum if that matches the characteristic impedance. In practice, it was a little bit high at 1/4 wavelength, and a bit low at 1/2 wavelength which would suggest the characteristic impedance might be a bit higher than 130 ohms. But 130 to 135 ohms is a reasonable range.

Note that I had to suspend the cable in the air, away from metal items or the walls, to get repeatable results with these measurements, although that shouldn’t be necessary with coax cable.

summary

So there are several different methods. I usually use at least two when possible to check my results, depending on the circumstances. Practice them using lines of known impedance, and see which one(s) seem to work best for your with the equipment you have available.